【作者】 栾军伟；

【导师】 刘世荣；

【作者基本信息】 中国林业科学研究院 ， 生态学， 2010， 博士

【摘要】 土壤呼吸是大气CO₂来源之一,在决定生态系统作为大气碳源汇方面起着重要作用。理解土壤呼吸调控机理对于我们更好的描述全球碳收支具有重要意义。但是,不同年龄段、不同森林类型间,土壤呼吸变异规律及其调控机理都存在不同程度的差异。本研究位于伏牛山地区宝天曼自然保护区,通过对该区主要森林类型之一锐齿栎林土壤呼吸的研究,旨在为计算该区森林生态系统碳收支提供理论基础。1)研究了暖温带锐齿栎林长期演替序列（40年生幼龄林（YO）、48年生中龄林（IO）、80年生成熟林（MO）、143年生过熟林（OGO））土壤呼吸（RS）时空变异性。四种林分土壤呼吸的时间变异很大程度上依赖于土壤5cm温度（T5）,T5揭示了土壤呼吸时间变异的73.8～82.5%。土壤含水量（SWC）仅与IO林土壤呼吸在时间上有弱相关,与其它林分土壤呼吸均无相关关系。四种林分土壤呼吸标准差（SD）呈明显的季节动态,与T5季节动态相一致,而土壤呼吸变异系数（CV）则无显著季节动态。土壤水分含量的空间变异系数与YO林和OGO林土壤呼吸空间变异系数呈显著线性相关关系。OGO林土壤呼吸变异性显著高于其它林分,因而,在95%置信区间内获得的土壤呼吸值控制在实际值的±20%误差范围内,OGO林需要取样点数高于其它林分。表层（0-10cm）土壤有机碳（SOC）,尤其是土壤轻组有机碳（LFOC）能很好的解释土壤累积呼吸量不同林龄间差异。土壤0-5cm总孔隙度（TP）与不同林龄间土壤累积呼吸量呈显著负相关关系,这种负相关关系可能来自于土壤毛管孔隙对土壤呼吸的限制或制约。随着森林演替,土壤活性碳储量及土壤物理性状等随之变化,继而影响了土壤表面碳通量,使土壤表面CO₂通量随林龄增加而增加。森林演替后期,土壤呼吸空间变异性明显增加。2)通过壕沟断根法结合根分解实验,区分估计了暖温带锐齿栎林长期演替序列（YO、IO、MO、OGO）生长季土壤自养（RR）和异养呼吸（RH）。并对其季节动态和时空变异调控因素进行了研究。研究发现,尽管根系呼吸与异养呼吸一样,在时间变异上可以很好的由土壤5cm温度通过指数模型解释,但是,YO和IO林根呼吸的峰值发生在9月份,较土壤温度的峰值滞后约30天。而且,与MO和OGO林不同,YO和IO林根系呼吸占总呼吸的比例在9月份出现了第二个峰值。2009年生长季（106-287天）累积自养和异养呼吸通量不同林龄间表现出显著差异。YO、IO、MO及OGO林生长季土壤异养呼吸累积值分别为431.72、452.02、484.62和678.93 g C m^-2,而其根呼吸累积值分别为191.94、206.51、321.13和153.03 g C m^-2。根系呼吸占土壤总呼吸的比例（RC）由YO林的30.78%增加到MO林的39.85%,然后降低到OGO林的18.39%。不同林龄间土壤0-10cm有机碳储量,尤其是活性有机碳储量（LFOC）与生长季累积土壤异养呼吸间呈显著相关。但是,细根生物量与根系呼吸间相关性不显著。异养呼吸的表面温度敏感性(Q₁₀)显著高于根系呼吸,两者分别为3.93和2.78。随林龄增加土壤毛管孔隙度显著降低,一定程度上解释了土壤呼吸不同林分间差异。该区锐齿栎长期演替序列土壤呼吸随林龄的增加而增加,主要由该区森林土壤异养呼吸较高贡献率所致。不同演替阶段,森林根系呼吸表现出不同的季节格局。本研究强调了土壤呼吸组分分离在评价林龄对土壤呼吸影响时的重要性。3)通过在两种林分分别建立40×60m固定样地,按10m网格划分后在网格交点测定土壤呼吸、土壤理化性状、根生物量及林分结构特征等,研究比较了暖温带地区相邻锐齿栎天然次生林（OF）与华山松人工林（PP）土壤呼吸时空变异性及其调控因素。测定从2008年10月至2009年10月。研究表明,整个生长季过程中,两种林分土壤呼吸空间格局均较为稳定。与锐齿栎次生林相比,华山松人工林土壤呼吸空间变异较小。各测定点土壤呼吸与5cm土壤温度均呈显著指数相关。研究发现,土壤呼吸与土壤水分含量（SWC）在空间上呈负相关,但是,水分饱和对气体扩散的限制作用并不是土壤水分含量较高区域呼吸量较低的主要原因。与土壤水分含量相比,土壤孔隙充水率（WFPS）能更好的解释土壤呼吸空间变异性。以土壤呼吸测量点为中心,半径4-5m范围内的胸高断面积和（BA）、最大胸径（max DBH）、平均胸径（mean DBH）等林分结构参数能很好的解释锐齿栎次生林土壤呼吸的空间变异性,而在华山松人工林中则未发现相似结果。多元逐步回归模型表明,土壤轻组有机碳（LFOC）含量、土壤持水力（WHC）共同解释了华山松人工林土壤呼吸空间变异的49.6%;而土壤持水力（WHC）、4米半径内最大胸径（max DBH4）、及土壤总孔隙（TP）共同解释了锐齿栎次生林土壤呼吸空间变异的64.2%。这表明生物与非生物因子在控制次生林与人工林土壤呼吸空间变异时的差异性。土壤呼吸温度敏感性(Q₁₀)与土壤碳库活度（LLFOC）、细根生物量（FR）在空间上呈正相关关系,而与土壤水分含量则呈负相关关系。森林覆盖变化造成土壤水分含量的显著降低可能是华山松人工林土壤呼吸温度敏感性显著增加的原因。本文研究结果强调了次生林与人工林土壤碳通量空间变异及其调控因子的差异性,对于精确估计区域碳收支具有重要意义。4)关于气候变暖对中国暖温带森林土壤碳氮循环潜在影响的研究尚未见报道,本研究将高海拔地区（1400m）锐齿栎林直径30cm、高40cm的原状土柱移至低海拔地区（620m）,将低海拔地区栓皮栎林原状土柱移至高海拔地区,通过土壤互置实验模拟气候变暖或变冷对土壤碳氮动态的影响。研究表明,低海拔样地土壤平均温度比高海拔样地约高3℃,而土壤水分含量约低6%。对控制（in situ）及移位处理（transfer）土柱土壤氮周转和土壤呼吸均进行了为期一年的观测。移位土壤净氮矿化率及净氮硝化化率比原位培养的分别高约254%和67%,移位土壤呼吸速率比原位培养的高约52%。在整个实验过程中,相对于较为稳定的增温效果（3℃）,土壤呼吸量对移位处理的响应程度则呈现明显的季节动态,在夏季时达到最大。土壤从高海拔到低海拔移位一年后,土壤微生物生物量碳显著降低,可溶性有机碳显著增加,土壤异养呼吸温度敏感性显著增加,而土壤异养基础呼吸明显降低。相反,土壤从低海拔到高海拔移位,对土壤碳、氮过程影响程度较低,各土壤呼吸参数及土壤活性有机碳含量均无显著变化。本研究表明,从高海拔到低海拔土壤移位短期内对土壤碳氮过程及土壤可用性底物均具有强烈的影响。更多还原

【Abstract】 Soil respiration is one of the atmosperic CO₂ sources, which plays an improtant role in determining whether an ecosystem is a carbon sink or source to the atmosphere. A good understanding of the machanisms underlying soil respiration will help us to better ascertain the global land carbon budget. However, the temporal and spatial variations of soil respiration and their controlling factors varied largely with forest succussional stages and types. This thesis was designed to examine soil respiration along with its controlling factors in oak forests at the Baotianman Natural Reserve in China in order to assess the regional carbon budget.1) The temporal and spatial variations of soil respiration （RS） was investigated in a warm-temperate oak chronosequence, in China. The oak choronosequence included a 40-year-old young oak forest （YO）, a 48-year-old intermediate oak forest （IO）, a 80-year-old mature oak forest （MO） and a 143-year-old old growth oak forest （OGO）. Temporal variations of RS of the four forests largely depended on soil temperature at 5cm depth （T5）, which explained 73.8～82.5% of the temporal variation of RS. In IO forest, soil water content （SWC） had a weak effect on the temporal variation of RS. The seasonal patterns of standard deviation （SD） of RS showed the similar trend with T5, while no seasonal trends of variation coefficients （CV） in RS was found for the four forests. In YO and MO forests, spatial variation coefficients of SWC correlated significantly positive with the spatial variation coefficients of RS. The spatial variation of RS was the highest in OGO among the oak chronosequence. Therefore, more sampling points were needed in OGO forest in order to obtain an average rate of RS within 20% of its actual value at the 95% confidence level. Top soil organic carbon （SOC）, especially soil light fraction organic carbon （LFOC） well explained the variation of cumulative RS among the stands. We found total porosity （TP） at 0-5cm soil depth correlated negatively with the cumulative RS, this may be due to the limitation of capillary porosity （CP） on RS. Soil labile carbon storage and soil physical properties varied with the forest succession, which influenced the CO₂ efflux. Furthermore, forest age had a positive effect on spatial varation of RS.2) Plot trenching and root decomposition experiments were conducted to partion soil respiration components in a warm-temperate oak chronosequence （YO, IO, MO and OGO forests） in China. Total soil surface CO₂ efflux （RS） was partitioned into rhizospheric （RR） and heterotrophic respiration （RH） across growing season of 2009. It was found that the temporal variations of RR and RH can be well explained by soil temperature at 5cm depth （T5） using exponential equation. However, RR of YO and IO forests peaked in September, while their T5 peaks advanced 30 days （in August）. Also, unlike MO and OGO forests, the contribution of RR to RS （RC） of YO and IO forests presented the second peak in September. There were significant differences in the cumulative RH and RR fluxes during the growing season among the four forests. The estimated RH values for YO, IO, MO and OGO forests averaged 431.72, 452.02, 484.62 and 678.93 g C m-2, respectively, while their corresponding RR averaged 191.94, 206.51, 321.13 and 153.03 g C m-2, respectively. The estimated RC increased from 30.78% in the YO forest to 39.85% in the MO forest and then declined to 18.39% in the OGO forest. There was significant correlation between soil organic carbon （SOC）, especially the labile organic carbon （LFOC） storage of 0-10cm soil depth and the cumulatve RH during the growing season. There was no significant relationship between RR and fine root biomass regardless of the stand age. Apparent temperature sensitivity (Q₁₀) of RH （3.93±0.27） is significantly higher than that of RR （2.78±0.73）. The capillary porosity decreased as the stand age increased, which accounted for the differences in cumulative RS among the four chronosequence. It was concluded that the positive effect of forest age on RS attributed mainly to the increasing proportion of RH with age. The seasonal patterns of RR varied with forest age. Our results emphasized the importance of respiration components partitioning when evaluating the age effect on soil respiration and its significance to future model construction.3) Factors that control spatiotemporal variations of soil respiration （RS） were assessed in a natural regenerated oak forest （OF） and a nearby pine plantation （PP） in warm-temperate area of China. RS, soil properties and stand structure were measured at 10m intervals in two 40×60m plots （35 grid points） for OF and PP from Oct. 2008 to Oct. 2009, respectively. The observed spatial pattern kept remarkably stable throughout the growing season. Compared to OF, PP showed relatively lower spatial variations of RS across the growing season. The spatial relationships between RS and soil water content （SWC） were found to be negative. However, the restriction of gas diffusivity in water-saturated soil was not the primary cause of the low RS in wetter regions. Compared to SWC, water filled pore space （WFPS） might be a better parameter to explain the spatial variation of RS. The stand structure parameters, such as basal area （BA）, max diameter at breast height （max DBH） and mean DBH within 4 or 5m of the measurement points accounted well for the spatial variation of RS in OF. However, no similar correlation was found in PP. Multilinear regression results showed that light fraction organic carbon （LFOC） and water hold capacity （WHC） explained 49.6% of the variation of RS in PP, while WHC, Max DBH（4） and total porosity （TP） explained 64.2% of the variation of RS in OF. This suggested that biotic and abiotic factors played different roles in controlling spatial variations of RS between OF and PP. Regardless of the stand, spatial distribution of carbon pool lability （LLFOC） and fine root biomass （FR） correlated positively with the spatial variation of apparent temperature sensitivity of RS (Q₁₀), while SWC negatively correlated with the spatial variation of Q₁₀. The significant higher Q₁₀ of PP compared to OF may due to the decreased SWC. Our findings ascertain the spatio-temporal variations of RS between plantation and naturally regenerated forests, which is useful to make an accurate estimation of regional carbon fluxes.4) Few research has been conducted on how climate change may affect the soil C and N processes of warm-temperate oak forest in China. Along the slope of the Funiu mountains, China, intact soil monoliths from a 1400m （Quercus acutidentata） oak forest were translocated to a 620m （Quercus variabilis） oak forest and vice versa. Through the intact soil monoliths reciprocal translocation experiment, the likely impacts of climate change on soil C and N processes were explored. The results showed the mean annual soil temperature at 5cm depth （T5） was about 3℃higher in the low-elevation site than that in the high-elevation site, while the soil water content （SWC） was about 6% lower in the low-elevation site than that in the high-elevation site. Net rates of N transformations and CO₂ fluxes were measured in high-elevation soil monoliths incubated in situ and soil monoliths transferred to the low-elevation site and vice versa. Net N mineralization and nitrification increased about 254% and 67% in transferred soil cores compared with in situ soil cores. Soil transfer significantly increased CO2 efflux （52%） compared to fluxes from soil monoliths incubated in situ. Soil transfer resulted in a relatively stable temperature increase and this warming effect on soil CO2 efflux increased as the weather get warmer. Soil microbial biomass carbon （MBC） decreased in transferred soil monoliths compared to in situ soil monoliths after one year incubation period （from 1.14 to 0.73）, while dissolved organic carbon （DOC） increased （from 0.23 to 0.27）. Furthermore, transferred soil monoliths from the high-elevation to the low-elevation raised soil respiration temperature sensitivity (Q₁₀) （from 2.73 to 3.42）, while reduced soil basal respiration （R0） （from 0.42 to 0.29） compared to soil monoliths incubated in situ. In contrast, transferred soil monoliths from the low-elevation to the high-elevation site （i.e. simulated global cooling） produced weak effects on soil C process as no significant changes in soil respiration parameters and soil labile carbon content were found. Our results suggested that short term soil translocation would lead to large impacts on soil C and N processes, and the soil substrate availability as well.更多还原